A Gaussian Process Based Model Predictive Controller for Nonlinear Systems with Uncertain Input-Output Delay

Abstract

In this paper, we propose a Model Predictive Controller (MPC) based on Gaussian process for nonlinear systems with uncertain delays and external Gaussian disturbances. We investigate the ability of Gaussian process based MPC on handling the variable delay that follows a Gaussian distribution through a properly selected observation horizon. To test the effectiveness of this approach, comparisons are made for the proposed Gaussian process based MPC and RBF (Radial Basis Function) neural networks by analyzing the time complexity and control performance. In simulations, two experiments are designed to verify the results of different systems, including a first-order nonlinear plant and a second-order nonlinear plant with variable delays and Gaussian noises. It is demonstrated that the proposed approach may achieve the desired results.